Here are some of the trends emerging from the conferences:
• Mobile WiMAX to dominate, even in "fixed" deployments
One common theme was the sense that fixed WiMAX will be increasingly sidelined by its mobile cousin. Most vendors now see 802.16d as a niche deployment, unlikely to gain global scale.
Instead, the mobile variant will be deployed even if the usage mode is essentially "fixed" (i.e. signal received via a roof antenna or fixed modem to a desktop). This is due both to the superior propagation qualities of OFDMA, and also to the improved scale economies gained by focusing on the mobile platform. We believe this proposition makes a great deal of sense – particularly given the incompatibility between the two flavors of the technology.
The timeline for delivery of certified mobile WiMAX equipment is now being pushed back to the end of 2007. While we heard quite a few comments to the effect that "2007 will be the year of mobile WiMAX" we believe it will, like 2006, be the year of pre-standard mobile WiMAX.
• Widespread availability of WiMAX-capable devices is critical
Discussions with Intel gave us a clear indication of how it will approach mobile WiMAX, and its prognosis is encouraging. It will offer mobile WiMAX PC cards in 2007 (mainly for evaluation purposes) before beginning to ship the Centrino WiMAX platform in late 2007 or early 2008. It believes that the "attach rate" will follow the path of the Wi-Fi-capable Centrino chipset, which grew from 10 percent of laptops shipped to over 80 percent in just four years.
Intel stated that there is considerable commonality in the Wi-Fi and WiMAX platforms, which will enable it to bring cost down rapidly. It is aiming for what it terms a "30-30" model, whereby customers should have to pay no more than $30 for the radio module, and no more than $30 for monthly data access.
• The embedded device model – as yet unproven
We heard several presentations in praise of the "embedded device" business model, whereby nearly every item of consumer electronics (and beyond) becomes the host of a WiMAX chip, enabling a "new paradigm" of productivity and convenience. This bears a resemblance to Japanese operator NTT DoCoMo’s mission statement at the height of the telecoms boom, when it totaled all refrigerators, dogs, cats and humans in Japan as potential bearers of cellular connectivity. That disappeared with the stock market downturn, but we’re now seeing the same story with a different technology.
We believe there are several obstacles to making this work as a paid-for service. First of all, many of the proposed consumer electronics (cameras, etc.) have neither the screen nor the interface to effectively manage a wireless connection. Secondly, to what extent will users be willing to pay for such a service? Success hinges on making this cheap and seamless with a customer’s existing billing relationship, but in practice this could involve an awkward co-existence of different user accounts and payment modes.
This said, the "one-to-many" business model is exciting. Particularly, we feel there is a window of opportunity in the three- to five-year timeframe for mobile WiMAX, when a considerable number of enabled devices will be in the market. This is also when there could be uncertainty around 3G Long Term Evolution (LTE), which could yet get mired in IPR-centered disagreement.
The 3G Partnership Project also has the challenge of managing the very different time scales of its members: NTT DoCoMo is issuing requests for LTE tender now, while some operators have barely deployed WCDMA.
This said, the embedded device model requires building a business system almost from scratch, persuading both consumer electronics vendors and customers of the relevance of such a service. This will prove extremely challenging.